Provision device, provision method and provision program

Abstract

To provide a provision device, a provision method and a provision program that can provide a transaction object considering an evaluation axis when a user selects a commodity.SOLUTION: A provision device 1 comprises: an estimation unit 132; a determination unit 133; and a provision unit 134. The estimation unit 132 estimates a plurality of evaluation axes for a target commodity to be selected and an importance degree of each evaluation axis on the basis of information of a commodity itself, user information of a user and a conversation process associated with a specific category which is performed with the user using a conversation type service. The determination unit 133 determines similarity among commodities using data similarity of each of a plurality of commodities corresponding to each of the plurality of evaluation axes, and the importance degree of each evaluation axis. The provision unit 134 provides a content in which a target commodity and a similar commodity being similar to the target commodity are arranged in an order depending on similarity among the commodities determined by the determination unit 133.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a providing apparatus, a providing method, and a providing program.

Background Art

[0002] Conventionally, an e-commerce street service that enables the purchase of trading targets sold by a plurality of merchants has been provided. As an example of such a service, a technique for proposing trading targets selected according to the attributes of users is known.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the above-described conventional technology, it may not be possible to say that trading targets are proposed in consideration of the evaluation axes when users select (evaluate) products.

[0005] The present application has been made in view of the above, and an object thereof is to provide trading targets in consideration of the evaluation axes when users select products and services including tangible and intangible items.

Means for Solving the Problems

[0006] The providing device according to the present application includes an estimation unit, a determination unit, and a providing unit. The estimation unit estimates a plurality of evaluation axes when a target product belonging to the specific category is selected and the importance of each evaluation axis based on a conversation process regarding the specific category conducted with the user using information on the product itself, user information of the user, and a conversational service. The determination unit determines the similarity between products using the similarity of data of each of the plurality of products belonging to the specific category corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis. The providing unit provides content in which the target product and similar products similar to the target product are arranged in an order according to the similarity between products determined by the determination unit.

Effect of the Invention

[0007] According to one aspect of the embodiment, it is possible to provide a transaction target in consideration of the evaluation axes when a user selects a product.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7A

Figure 7B

Figure 7C

Figure 7D

Figure 7E

Figure 8

Figure 9

[0009] Hereinafter, embodiments for implementing the provision device, provision method, and provision program according to the present application (hereinafter referred to as "embodiments") will be described in detail with reference to the drawings. Note that the provision device, provision method, and provision program according to the present application are not limited by this embodiment. Also, each embodiment can be appropriately combined within a range that does not conflict with the processing content. In the following embodiments, the same parts are denoted by the same reference numerals, and redundant explanations are omitted.

[0010] 〔1. An example of the provision process shown by the provision device〕 First, the provision process executed by the provision device 1 according to the embodiment will be described. FIG. 1 is an explanatory diagram showing an example of the provision process according to the embodiment. Here, the provision device 1 uses conversational search to estimate the evaluation axes when a product that the user is interested in is searched (selected) from the conversation process with the user and the information of the product itself, and provides the user with other products whose values corresponding to the estimated evaluation axes are similar. The conversational search here refers to, for example, a conversational search service including a conversational service using AI such as ChatGPT (Generative Pre-trained Transformer). The conversation process here refers to a group of conversations obtained from such a conversational search service.

[0011] Hereinafter, with reference to FIG. 1, an example of the provision process executed by the provision device 1 will be described. Here, it is assumed that the category of the product that the user is interested in is a device for navigating a vehicle (abbreviated as "car navigation").

[0012] As shown in FIG. 1, the provision device 1 receives (S1) a conversation process C1 conducted with the user terminal 10 regarding products in the category from the user terminal 10. For example, the provision device 1 receives a question regarding a product in the category input from the user terminal 10 using a conversational search service. The provision device 1 outputs an answer to the question to the user terminal 10. That is, the provision device 1 repeats the conversation with the user terminal 10 using the conversational search service. Here, the user terminal 10 sequentially receives questions such as "Tell me the features of ●● car navigation" and "What is the appropriate screen size?" input by the user in the text box. Then, the provision device 1 outputs an answer selected by a dialog service (e.g., GPT) to the question to the user terminal 10. That is, the provision device 1 receives the conversation process C1 conducted with the user terminal 10 from the user terminal 10.

[0013] Then, based on the conversation traced up to just before the conversation process C1, the provision device 1 outputs (S2) the content of the recommended category of products to the user terminal 10. For example, the provision device 1 selects a recommended category of products using the conversational search service based on the conversation in the conversation process C1 and outputs the content of the selected product to the user terminal 10.

[0014] Subsequently, the provision device 1 receives (S3) an instruction from the user terminal 10 to select other candidates from the conversation process C1 so far.

[0015] Then, the providing device 1 estimates a plurality of evaluation axes and the importance of each evaluation axis from the information of the product itself, the user information of the user, and the conversation process C1 with the user (S4). Here, the user information of the user refers to, for example, information such as the user's usage history (purchase history) and search history in various services, and is collected from the user terminal 10 or other server devices. For example, the providing device 1 uses the conversation process C1, the user information of the user, the information of the product itself, and a model that has been trained to generate answers to the input questions to estimate a plurality of evaluation axes and the importance of each evaluation axis. Here, the information of the product itself refers to the data associated with the product. When the category of the product is "car navigation", it includes the screen size, accuracy, manufacturer, price, etc. of the product. The information of the product itself can be obtained, for example, by using a pre-set product information database. Taking a specific example, the providing device 1 inputs a predetermined instruction text (prompt) defined in advance, together with the conversation process C1 and the user information of the user, into the model to estimate the evaluation criteria and scores. Here, the predetermined instruction text (prompt) is, for example, the text "Please extract the evaluation axes that are considered important, useful, and effective in the process of conversing with me and selecting a recommended product so far. Next, with the extracted evaluation axes on the vertical axis and the product selected as the recommended product at the top of the horizontal axis, select similar products on the evaluation axis from the product information database and present them in a table. Regarding the determination of product similarity, please understand 'above', 'below', 'better', 'worse' and conform to the user's intention." The prompt is not limited to this.

[0016] The prompts mentioned here include candidates for evaluation axes that are preset for each product category, and may also include sentences that allow selection from among these candidates. For example, it may include sentences such as "Note that the evaluation axis should be selected from the following candidates. Screen size, number of captured satellites, user evaluation, TV compatibility." That is, the providing device 1 may include the preset candidates in the prompt according to the product category of the candidates to be provided to the user.

[0017] Also, the user information input to the model is not limited to user information such as the user's usage history (purchase history) and search history, and information corresponding to the product category may be selected. For example, the providing device 1 uses a model that estimates the relevance between the product category and the content of the conversation or a calculation method for vector similarity to select information with a high degree of relevance to the product category to be proposed from among the user information, and may include the selected information as the user information. Thereby, the providing device 1 can reduce the entire prompt and prevent the deterioration of the estimation accuracy caused by using user information with low relevance.

[0018] In this way, the providing device 1 may control the candidates for values and user information as the content of the prompt according to the product category. In addition, the providing device 1 may appropriately select as the content of the prompt according to, in addition to the product category, click history, past viewed products, conversations with the user, and the like.

[0019] Note that the model is a model that is learned to output a response sentence corresponding to the input question sentence, and is a language model that performs natural language processing such as GPT or BERT (Bidirectional Encoder Representations from Transformers). The model is within the providing device 1 and is independently created by the operator who manages the providing device 1. It is desirable that the input information be learned so as not to be used as a new response to conceal information such as the input personal information.

[0020] Here, the providing device 1 assumes that the evaluation axes of the category are the screen size, the number of captured satellites, the user evaluation, and the TV compatibility. The providing device 1 assumes that the importance levels of the evaluation axes, in descending order, are the screen size, the number of captured satellites, the user evaluation, and the TV compatibility.

[0021] Subsequently, the providing device 1 determines the similarity between products using a plurality of evaluation axes and the importance level of each evaluation axis (S5). For example, for a recommended product, the providing device 1 acquires respective values corresponding to the plurality of evaluation axes from a preset product information database. Also, for the product group of the category, the providing device 1 acquires respective values corresponding to the plurality of evaluation axes from the product information database. Then, the providing device 1 determines the similarity between the recommended product and each product belonging to the category such that the higher the importance level of the evaluation axis to which the corresponding value is similar, the higher the ranking.

[0022] Note that the providing device 1 considers "above", "below", "better", and "worse" based on the content of the conversation with the user so as to conform to the user's intention regarding the determination of the similarity between the recommended product and each product belonging to the category based on the instruction text (prompt). As an example, when "screen size" is assumed as an evaluation axis, the providing device 1 considers whether "larger" or "smaller" is better based on the content of the conversation and in accordance with the user's intention, and determines the similarity between the recommended product and each product belonging to the category.

[0023] Then, the providing device 1 generates content arranging the recommended product and similar products similar to the recommended product in an order according to the similarity between the products (S6). For example, for similar products similar to the recommended product, the providing device 1 generates content arranging the recommended product and similar products similar to the recommended product with the higher the similarity of the value corresponding to the evaluation axis with higher importance level, the higher the ranking.

[0024] Then, the providing device 1 provides the generated content to the user terminal 10 (S7). For example, the providing device 1 arranges a plurality of evaluation axes on the vertical axis, arranges recommended products and similar products on the horizontal axis, and provides the user terminal 10 with Table C2 in which values corresponding to each evaluation axis for each product are set for the arranged content. Each evaluation axis may be arranged in the order of importance. Thereby, the providing device 1 can provide the user terminal 10 with products belonging to the category desired by the user in consideration of the evaluation axes when the user selects a product.

[0025] Note that the embodiment is not necessarily limited to the above example.

[0026] As an example, the providing device 1 may be configured to estimate evaluation axes that have not yet been estimated other than the evaluation axes estimated from the conversation process according to the user's request. For example, the providing device 1 may estimate an evaluation axis that has not appeared in the conversation with the user and is specific to the category according to the setting of the radio button performed by the user. In the embodiment, radio button B1 is shown as an example of the radio button. When radio button B1 is set to ON, "Bluetooth (registered trademark) compatible" is estimated as an evaluation axis specific to the car navigation that has not appeared in the conversation with the user. Thereby, the providing device 1 can add evaluation axes that the user himself / herself has not noticed and enable the user to appropriately select a product.

[0027] Also, as another example, in addition to the similarity between products, the providing device 1 may generate content in which recommended products and similar products similar to the recommended products are arranged in an order according to the advertising fees corresponding to the pre-set products, and provide the content to the user terminal 10. Thereby, the providing device 1 can increase the priority of similar products for which a high advertising fee is paid by the operator and provide them to the user terminal 10.

[0028] Also, as another example, in addition to the similarity between products, the providing device 1 may generate content that arranges recommended products and similar products similar to the recommended products in an order according to the click-through rate of users including other users to similar products similar to the recommended products, and provide it to the user terminal 10. The click-through rate refers to, for example, the ratio of the provided similar products that are clicked by the user. Thereby, the providing device 1 can consider the probability that the click-through rate is high, that is, the probability that the user's interest is likely to be high, and provide it to the user terminal 10.

[0029] Also, as another example, in addition to the similarity between products, the providing device 1 may generate content that arranges recommended products and similar products similar to the recommended products in an order according to the multiplication of the advertising fee corresponding to a preset product and the click-through rate of the posted product, and provide it to the user terminal 10. Thereby, the providing device 1 can not only increase the priority of similar products for which a high advertising fee is paid by the business operator, but also consider the probability that the user clicks and provide it to the user terminal 10.

[0030] Also, as another example, in addition to the similarity between products, the providing device 1 may generate content that arranges recommended products and similar products similar to the recommended products in an order according to the multiplication of the advertising fee corresponding to a preset product, the click-through rate of the posted product, and further the lifetime value of using the product, and provide it to the user terminal 10. Thereby, the providing device 1 can not only increase the priority of similar products for which a high advertising fee is paid by the business operator, but also consider the probability that the user clicks and the subsequent usage value, and provide it to the user terminal 10.

[0031] As another example, the providing device 1 may provide comparison content regarding the arrangement of each evaluation axis so that, beside each evaluation axis, other users such as experts related to the category of the product selected by the user or other users followed by the user himself / herself can see the evaluation axes that are emphasized when selecting products belonging to the category. Thereby, when the user selects a product belonging to the category, the providing device 1 can provide the user with useful information that is not trapped within the user's framework.

[0032] As another example, the providing device 1 may change the importance levels of a plurality of evaluation axes according to the click-through rate of other users on similar products similar to the recommended product. For example, for a similar product that has low similarity to the recommended product but has a high click-through rate, if the value of "screen size", which indicates the evaluation axis with the highest current importance, is smaller than that of the recommended product, but the value of "user evaluation", which indicates the evaluation axis with the middle current importance, is higher than that of the recommended product, the providing device 1 may change the importance of "user evaluation" to be higher than that of "screen size" for the evaluation axis. Thereby, the providing device 1 can change the importance levels of the evaluation axes according to the actual access of the user to similar products.

[0033] As another example, although the providing device 1 has been described with respect to products, it may also be directed to services or things that cannot be fully grasped as products. For example, it may be applied to travel / accommodation, real estate, movies / music / books, finance, restaurants, recipes, job hunting / career change, further education, beauty salons, etc. When the target is travel / accommodation, possible evaluation axes may include price, ranking (rating), surface time, region, region type, building type, user review score, etc. When the target is real estate, possible evaluation axes may include price, ranking (rating), required time, age of construction, size, floor plan, etc. When the target is job hunting / career change, possible evaluation axes may include salary level (hourly wage level), turnover rate (length of service), welfare benefits (various allowances), region, industry type, job type, etc. When the target is further education, possible evaluation axes may include faculty, academic system, employment record, deviation value, region, etc. When the target is a beauty salon, possible evaluation axes may include price, evaluation of technique (number of ☆ or favorite hairstyles), region, celebrities who go there, etc.

[0034] [2. Configuration of the Providing System] Next, with reference to FIG. 2, the configuration of the providing system 100 according to the embodiment will be described. FIG. 2 is an explanatory diagram showing an example of the configuration of the providing system according to the embodiment. As shown in FIG. 2, the providing system 100 according to the embodiment includes a providing device 1 and a user terminal 10. The providing device 1 and the user terminal 10 are communicably connected by wire or wirelessly via a network N. Note that the providing system 100 may include the user terminal 10.

[0035] The user terminal 10 according to the embodiment is an information processing device used by a user who accesses content such as a web page displayed on a browser or content for an application. For example, the user terminal 10 is a desktop PC (Personal Computer), a notebook PC, a tablet terminal, a mobile phone, a PDA (Personal Digital Assistant), or the like.

[0036] [3. Configuration of the Providing Device] The providing device 1 according to the embodiment is realized by, for example, a server device or the like. As shown in FIG. 2, the providing device 1 includes a communication unit 11, a storage unit 12, and a control unit 13.

[0037] The communication unit 11 is realized by, for example, a NIC (Network Interface Card) or the like. Then, the communication unit 11 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the user terminal 10.

[0038] The storage unit 12 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. As shown in FIG. 2, the storage unit 12 includes a user information database 121, an estimated information database 122, a product information database 123, and a model database 124.

[0039] The user information database 121 stores various types of information about the user. Here, an example of the information stored in the user information database 121 will be described with reference to FIG. 3. FIG. 3 is a diagram showing an example of the user information database. In the example of FIG. 3, the user information database 121 stores items such as "attribute information", "search history", "purchase history", "posting information", "review information", "value information", and "purchase tendency" in association with the "user ID".

[0040] "User ID" indicates identification information for identifying a user. "Attribute information" indicates the attributes of a user (demographic attributes and psychographic attributes). "Search history" indicates the search history of a user, and stores information such as, for example, a search query and the date and time when the search query was entered. "Purchase history" indicates the purchase history of products by a user, and stores information such as, for example, information indicating the purchased product, the category of the product, and the date and time of purchase. "Posting information" indicates information posted by a user to a predetermined service, and stores information such as, for example, the posted text information and the date and time of posting. "Review information" indicates the review posted by a user for a product, and stores information such as, for example, the product that is the subject of the review, the text information indicating the review, and the date and time when the review was posted. "Value information" indicates the values of a user regarding the purchase of products (for example, emphasizing low price or simplicity). "Purchase tendency" indicates the tendency of a user to purchase products, estimated based on information such as the search history, posting information, and purchase history.

[0041] That is, in FIG. 3, an example is shown in which the attribute information of the user identified by the user ID "UID#1" is "Attribute Information #1", the search history is "Search History #1", the purchase history is "Purchase History #1", the posting information is "Posting Information #1", the review information is "Review Information #1", the value information is "Value Information #1", and the purchase tendency is "Purchase Tendency #1".

[0042] Returning to FIG. 2, the estimation information database 122 stores various types of information related to estimation. Here, using FIG. 4, an example of the information stored in the estimation information database 122 will be described. FIG. 4 is a diagram showing an example of the estimation information database. In the example of FIG. 4, the estimation information database 122 has items such as "User ID", "Category", "Evaluation axis information", and "Score information".

[0043] "User ID" indicates identification information for identifying a user. "Category" indicates the category of the product specified by the user. "Evaluation axis information" indicates the evaluation axis of the category of the product specified by the user. "Score information" indicates the score of the evaluation axis of the category of the product specified by the user. Note that multiple evaluation axes are stored in the evaluation axis information of the category of the product corresponding to each individual user ID. And in the score information corresponding to the evaluation axis information, scores indicating each important element of each evaluation axis are stored.

[0044] That is, in FIG. 4, as an example, the category specified by the user identified by the user ID "UID#1" is "Category #1", and an example is shown where the evaluation axis information of the category is "Evaluation Axis Information #11" and the score information is "Score Information #11".

[0045] Returning to FIG. 2, the product information database 123 stores various information about products. Here, using FIG. 5, an example of the information stored in the product information database 123 will be described. FIG. 5 is a diagram showing an example of the product information database. In the example of FIG. 5, the product information database 123 stores items such as "data associated with the product", "advertising fee", "click-through rate", etc. in association with the "product name". Note that the product information database 123 may also store items such as "click unit price" obtained by multiplying the "advertising fee" and the "click-through rate", and "lifetime value" in association with the "product name".

[0046] "Product name" indicates the name for identifying the product. "Category" indicates the category of the product. "Data associated with the product" indicates various data associated with the product. As an example, when the category of the product is "car navigation", the "data associated with the product" includes "7 inches" as the screen size and "3 units" as the number of satellites captured. "Advertising fee" indicates the price bid for the business operator to publish the advertisement of the product. The click-through rate indicates the ratio of the users clicking on the provided product.

[0047] That is, in FIG. 5, as an example, when the product name is "Product Name #1", an example is shown where "Data #1" is the data associated with the product, "Advertising Fee #1" is the advertising fee, and "Click #1" is the click rate.

[0048] Returning to FIG. 2, the model database 124 stores a model that has been trained to generate an answer to the input question.

[0049] The control unit 13 is a controller, which is realized, for example, by various programs stored in the storage device inside the providing device 1 being executed with the RAM as the working area by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like. Also, the control unit 13 is a controller, which is realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). As shown in FIG. 2, the control unit 13 according to the embodiment has a reception unit 131, an estimation unit 132, a determination unit 133, and a provision unit 134, and realizes or executes the functions and operations of the provision process described below.

[0050] The reception unit 131 receives a conversation process using the interactive search service from the user terminal 10. For example, in the example of FIG. 1, the reception unit 131 receives a conversation process C1 about a specific category from the user terminal 10, and outputs the content of the recommended products belonging to the specific category to the user terminal 10 based on the conversation followed up to just before the conversation process C1.

[0051] The estimation unit 132 estimates a plurality of evaluation axes and the importance of each evaluation axis when a recommended product belonging to a specific category is selected based on the information of the product itself, the user information of the user, and the conversation process. For the information of the product itself, the data associated with the product in the product information database 123 may be used.

[0052] In the example of FIG. 1, the estimation unit 132 receives an instruction from the user terminal 10 to select other candidates different from the recommended products from the previous conversation process C1. Then, the estimation unit 132 acquires user information including the purchase history and search history corresponding to the target user ID from the user information database 121.

[0053] Then, the estimation unit 132 uses the received conversation process C1, the acquired user information of the user, the product information database 123, and the model database 124 to estimate a plurality of evaluation axes and the importance of each evaluation axis. Specifically, the estimation unit 132 inputs an instruction sentence (prompt) indicating that it is to output, based on the conversation process C1, the user's user information, and the product information database 123, the evaluation axes considered by the user when selecting products in the recommended category and the scores indicating the importance of the evaluation axes in the category, together with the conversation process C1 and the user's user information, into the model, thereby estimating the evaluation axes and the scores indicating the importance of the evaluation axes. Then, the estimation unit 132 associates the estimated evaluation axes and the scores indicating the importance of the evaluation axes with the user ID and the category and stores them in the estimation information database 122. Note that the instruction sentence (prompt) is a predefined instruction sentence. In the example of FIG. 1, the instruction sentence (prompt) is the sentence "Please extract the evaluation axes that are considered important, useful, and effective in the process of conversing with me and selecting the recommended products. Next, with the extracted evaluation axes on the vertical axis and the products selected as the recommended products at the top of the horizontal axis, select similar products from the product information database on the evaluation axes and present them in a table. Regarding the determination of product similarity, please understand 'above', 'below', 'better than', 'worse than' and conform to the user's intention."

[0054] The determination unit 133 determines the similarity between products by using the data associated with each product corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis. In the example of FIG. 1, the determination unit 133 acquires, from the recommended product information database 123, the values corresponding to each of the plurality of evaluation axes corresponding to the product name of the recommended product. Then, for the group of products belonging to the category, the determination unit 133 acquires, from the product information database 123, the values corresponding to each of the plurality of evaluation axes corresponding to each product name. Then, the determination unit 133 determines the similarity between the recommended product and each product belonging to the category such that the higher the importance of the evaluation axis, the closer the values corresponding to the evaluation axis are, and the higher the ranking.

[0055] The providing unit 134 provides the content in which the target product and the similar products similar to the target product are arranged in the order according to the similarity between the products. In the example of FIG. 1, for the similar products similar to the recommended product, the providing unit 134 generates the content in which the recommended product and the similar products similar to the recommended product are arranged with the higher the similarity of the values corresponding to the evaluation axes with higher importance, the higher the ranking. Then, the providing unit 134 provides the user terminal 10 with the content in which a plurality of evaluation axes are arranged on the vertical axis, the recommended product and the similar products are arranged on the horizontal axis, and Table C2 in which the values associated with each product for each evaluation axis are arranged. When arranging a plurality of evaluation axes on the vertical axis, the providing unit 134 may arrange the plurality of evaluation axes in the order of the importance of each evaluation axis. That is, the providing unit 134 may arrange each evaluation axis so that the higher the importance, the higher the position.

[0056] Note that although the providing unit 134 has been described as generating the content in which the target product and the similar products similar to the target product are arranged when providing, it is not limited thereto.

[0057] As an example, in addition to the similarity between products, the providing unit 134 may generate content that arranges a target product and similar products similar to the target product in an order according to the advertising fee corresponding to a preset product. For example, the providing unit 134 refers to the product information database 123 to obtain the advertising fee corresponding to a similar product similar to the recommended product. Then, the providing unit 134 rearranges the similar products with the products having a higher advertising fee prioritized, and generates content that arranges the recommended product and similar products similar to the recommended product. Thereby, the providing unit 134 can increase the priority of similar products for which a high advertising fee is paid to the business operator and provide it to the user terminal 10.

[0058] Also, as another example, in addition to the advertising fee, the providing unit 134 may generate content that arranges a target product and similar products similar to the target product in an order according to the click-through rate of users including other users corresponding to a preset product. For example, the providing unit 134 refers to the product information database 123 to obtain the click-through rate corresponding to a similar product similar to the recommended product. Then, the providing unit 134 rearranges the similar products with the products having a higher click-through rate prioritized, and generates content that arranges the recommended product and similar products similar to the recommended product. Thereby, the providing unit 134 can increase the priority of similar products with a high click-through rate and provide it to the user terminal 10.

[0059] Also, as another example, the providing unit 134 may generate content in which a target product and similar products similar to the target product are arranged in an order corresponding to the multiplication of an advertising fee corresponding to a preset product and the click-through rate of the product. For example, the providing unit 134 refers to the product information database 123 to obtain the advertising fee and click-through rate corresponding to similar products similar to the recommended product. Then, the providing unit 134 rearranges the similar products with priority given to the products with higher values obtained by multiplying the advertising fee and the click-through rate, and generates content in which the recommended product and similar products similar to the recommended product are arranged. Thereby, the providing unit 134 can increase the priority of similar products with a high click unit price and provide it to the user terminal 10.

[0060] Also, as another example, the providing unit 134 may generate content in which a target product and similar products similar to the target product are arranged in an order corresponding to the multiplication of an advertising fee corresponding to a preset product, the click-through rate of the product, and the lifetime value of using the product. For example, the providing unit 134 refers to the product information database 123 to obtain the advertising fee, click-through rate, and lifetime value corresponding to similar products similar to the recommended product. Then, the providing unit 134 rearranges the similar products with priority given to the products with higher values obtained by multiplying the advertising fee, the click-through rate, and the lifetime value, and generates content in which the recommended product and similar products similar to the recommended product are arranged. Thereby, the providing unit 134 can increase the priority of similar products so as to increase the profit on the merchant side and provide it to the user terminal 10.

[0061] As another example, when providing, the providing unit 134 may further arrange comparison content beside the evaluation axes that other users consider important among the multiple evaluation axes of the user. For example, the providing unit 134 arranges and provides comparison content beside the multiple evaluation axes arranged on the vertical axis so that it is possible to know the evaluation axes that other users consider important when selecting the target product. Here, the other users mentioned here may be other users different from the user, other users of the same age as the user, or other users who are experts regarding the category.

[0062] As another example, the providing unit 134 may convert the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis into various formats and provide them. For example, the providing unit 134 may represent the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis in tabular form. Also, the providing unit 134 may represent the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis in tabular form, and may also represent the differences from the recommended products by emphasizing them, or may represent the similarity indicating how similar each product is to the recommended product. Further, the providing unit 134 may display radio buttons or scroll bars so as to be able to estimate evaluation axes that are not mentioned in the conversation with the user and are specific to the category. In the example of FIG. 1, the radio buttons correspond to the buttons represented by reference sign B1 and have ON for prompting estimation and OFF for not prompting estimation. The scroll bar has a bar that can visually represent how many minutes to perform the estimation. Also, the providing unit 134 may convert the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis into star marks of n-level evaluation and represent them in tabular form. Further, the providing unit 134 may represent the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis in a polygonal chart corresponding to the number of evaluation axes.

[0063] [4. An Example of Provision] Here, an example of the provision according to the embodiment will be described with reference to FIG. 6. FIG. 6 is a diagram showing an example of the provision according to the embodiment. An example of the presentation shown in FIG. 6 is a case where the providing unit 134 provides, in tabular form, values corresponding to each evaluation axis corresponding to the recommended product g0 and similar products g1, g2, ··· arranged on the horizontal axis. On the horizontal axis, the contents of the recommended product g0 and similar products g1, g2, ··· similar to the recommended product g0 are shown. Also, a plurality of evaluation axes are shown on the vertical axis. Here, the evaluation axes include screen size, number of captured satellites, user evaluation, TV compatibility, and Bluetooth compatibility.

[0064] The similar products g1, g2, ··· on the horizontal axis are arranged in ascending order, for example, in accordance with the degree to which the values corresponding to the evaluation axes with high importance are similar to those of the recommended product g0. Each evaluation axis on the vertical axis is shown higher up as its importance is higher. And for each product, the values corresponding to each evaluation axis are represented. Here, since the similar product g1 has values corresponding to evaluation axes with higher importance that are more similar to those of the recommended product g0 than the similar product g2, it is arranged closer to the recommended product g0.

[0065] [5. Another Example of Provision] Also, another example of the provision according to the embodiment will be described with reference to FIGS. 7A to 7E. FIGS. 7A to 7E are diagrams showing another example of the provision according to the embodiment.

[0066] An example of the presentation shown in FIG. 7A is that the providing unit 134 represents, in tabular form, the values corresponding to each evaluation axis corresponding to the recommended product g0 and similar products g1, g2, ··· arranged on the horizontal axis as shown in FIG. 6. And the providing unit 134 provides, for example, by emphasizing the differences between each similar product g1, g2, ··· and the recommended product g0. Further, the providing unit 134 provides a similarity degree indicating how similar each similar product g1, g2, ··· is to the recommended product g0.

[0067] Here, for example, the similar product g1 differs from the recommended product in terms of screen size, user evaluation, and Bluetooth compatibility. Therefore, the providing unit 134 represents the different values of screen size, user evaluation, and Bluetooth compatibility using underlines. Furthermore, the providing unit 134 represents "95% match", indicating that the similar product g1 is 95% similar to the recommended product g0.

[0068] An example of the presentation shown in FIG. 7B is that the providing unit 134 represents, in tabular form, the values corresponding to each evaluation axis associated with the recommended product g0 and similar products g1, g2,... arranged on the horizontal axis as shown in FIG. 6. And the providing unit 134 is a case where a radio button B1 and a scroll bar B2 are displayed so as to be able to estimate an evaluation axis that has not appeared in the conversation with the user and is specific to the category.

[0069] Here, on the left side of the table, a radio button B1 and a scroll bar B2 are shown. The radio button B1 has ON for prompting estimation and OFF for not prompting estimation for the evaluation axis, and ON for prompting estimation is shown. Also, the scroll bar B2 has a bar that can visually represent how many minutes of estimation are to be performed for the evaluation axis, and the number of estimations can be increased as it is scrolled to the right. In FIG. 7B, in addition to screen size, number of captured satellites, user evaluation, TV compatibility, and Bluetooth compatibility as evaluation axes, the production country and map update frequency are added according to the scroll position of the scroll bar B2.

[0070] An example of the presentation shown in FIG. 7C is a case where the providing unit 134 converts the values corresponding to each evaluation axis associated with the recommended product and similar products arranged on the horizontal axis into star marks of n - stage evaluation and represents them. Here, it is a case where image quality, accuracy, sound quality, ease of use, and price are estimated for the evaluation axis. Here, the values corresponding to each evaluation axis of each product are represented by converting them into star marks of 5 - stage evaluation.

[0071] In the case of such a provision, the estimation unit 132 may input the following instruction text (prompt) into the model even when the input is as follows. The prompt may be, for example, "Please extract the evaluation axes that you think are important, useful, and effective in the process of conversing with me and selecting recommended products so far. Next, with the extracted evaluation axes on the vertical axis and the recommended product selected as the leading product on the horizontal axis, select products similar on the evaluation axes from the product information database and present them in a table. In the table, represent them with star ratings referring to the databases and websites consulted."

[0072] An example of the presentation shown in FIG. 7D is a case where the providing unit 134 represents the values corresponding to each evaluation axis corresponding to the recommended products and similar products arranged on the horizontal axis in a polygonal chart for the number of evaluation axes. Here, it is a case where the plurality of evaluation axes and the values corresponding to each evaluation axis shown in FIG. 7C are represented in a polygonal chart. In the chart of similar products, in addition to its own chart, the chart of the recommended product is faintly shown.

[0073] In the case of such a provision, the estimation unit 132 may input the following instruction text (prompt) into the model even when the input is as follows. The prompt may be, for example, "Please extract the evaluation axes that you think are important, useful, and effective in the process of conversing with me and selecting recommended products so far. Next, with the recommended product selected as the leading product, select products similar on the extracted evaluation axes from the product information database and represent them in a polygonal chart graph with the extracted evaluation axes as vertices." In addition, the prompt may be added with "Please rate the score of the evaluation axis for each product in each chart graph on a scale of 5 points referring to the databases and websites consulted." and "Please represent the score of the chart graph of the recommended product in a faint polygonal chart graph so that it can be compared in the chart graph of similar products."

[0074] An example of the provision shown in FIG. 7E is a case where the provision unit 134 arranges comparison content beside the evaluation axes that other users consider important among a plurality of evaluation axes of the user when providing. For example, the provision unit 134 arranges comparison content beside the plurality of evaluation axes arranged on the vertical axis so that it is possible to know the evaluation axes that other users considered important when selecting products belonging to the target category. When arranging the comparison content, the provision unit 134 may represent the comparison content with different colors and shapes according to the degree of importance. As an example, comparison content with a gold crown pattern, a silver crown pattern, and a copper crown pattern may be represented in descending order of importance. Further, the provision unit 134 may arrange a sorting button above the comparison content, and the evaluation axes may be represented in ascending / descending order by the sorting button.

[0075] Here, in the left figure, the comparison content is represented with colors changed according to the degree of importance so that it is possible to know the evaluation axes that other users considered important when selecting products in the category as the "evaluation axes of everyone". Also, in the middle figure, the comparison content is represented with colors changed according to the degree of importance so that it is possible to know the evaluation axes that users of the same age as the user considered important when selecting products in the category as the "evaluation axes of the same age as you". Further, in the right figure, the comparison content is represented so that it is possible to know the evaluation axes that other users who are experts in the category of the product selected by the user considered important when selecting products in the category as the "evaluation axes emphasized by experts".

[0076] [6. Flow of Provision Processing] Next, with reference to FIG. 8, the procedure of the provision process of the provision device 1 according to the embodiment will be described. FIG. 8 is a flowchart showing an example of the procedure of the provision process according to the embodiment. It is assumed that the user using the user terminal 10 wants to purchase a product belonging to a specific category.

[0077] As shown in FIG. 8, the provision device 1 outputs the content of the recommended product belonging to a specific category to the user terminal 10 based on the conversation process conducted with the user terminal 10 (step S11).

[0078] Subsequently, the providing device 1 determines whether it has received an instruction from the user terminal 10 to select another candidate (step S12). If it is determined that the instruction to select another candidate has not been received from the user terminal 10 (step S12; No), the providing device 1 repeats the determination process until the instruction is received.

[0079] On the other hand, if it is determined that an instruction to select another candidate has been received from the user terminal 10 (step S12; Yes), the providing device 1 estimates a plurality of evaluation axes and the importance of each evaluation axis from the information of the product itself, the user information of the user, and the conversation process (step S13). For example, when selecting a recommended product belonging to a specific category, the providing device 1 inputs an instruction sentence (prompt) indicating that it is instructed to output, based on the conversation process, the user's user information, and the product information database 123, an evaluation axis that the user considers and a score indicating the importance of the evaluation axis in the category, into the model together with the conversation process and the user's user information, thereby estimating the evaluation axis and the score indicating the importance of the evaluation axis. Note that the user information of the user may be acquired from the user information database 121.

[0080] Then, the providing device 1 determines the similarity between the recommended product and other products using the plurality of estimated evaluation axes and the importance of each evaluation axis (step S14). For example, the providing device 1 acquires values corresponding to each of the plurality of evaluation axes corresponding to the product name of the recommended product from the product information database 123. Then, for the product group belonging to the category, the providing device 1 acquires values corresponding to each of the plurality of evaluation axes corresponding to each product name from the product information database 123. Then, between the recommended product and the product group belonging to the category, as an example, the providing device 1 determines the similarity between the recommended product and each product belonging to the category such that the higher the importance of the evaluation axis, the more similar the corresponding values are, and the higher the ranking.

[0081] Then, the providing device 1 provides the content of the recommended product and the similar products to the user terminal 10 in an order according to the similarity between products (step S15).

[0082] [7. Modification Example] In the above-described embodiment, the reception unit 131 receives the conversation process using the interactive search service from the user terminal 10 and outputs the content of the recommended products belonging to a specific category to the user terminal 10. Then, the estimation unit 132 was described as estimating a plurality of evaluation axes when a recommended product belonging to a specific category is selected and the importance of each evaluation axis based on the information of the product itself, the user information of the user, and the conversation process. However, the above-described embodiment is merely an example, and various modifications and applications are possible.

[0083] For example, the reception unit 131 may directly receive the product name from the user terminal 10. In such a case, the estimation unit 132 estimates a plurality of evaluation axes corresponding to the characteristics of the product with the received product name and the importance of each evaluation axis based on the information of the product itself and the user information of the user. As the information of the product itself, the data associated with the product in the product information database 123 may be used.

[0084] As an example, the estimation unit 132 estimates a plurality of evaluation axes and the importance of each evaluation axis by using the received product name, the user information of the user, the product information database 123, and the model database 124. The user information of the user may be obtained from the user information database 121 and include the purchase history and search history corresponding to the target user ID. Specifically, when selecting a product in a recommended category, the estimation unit 132 inputs an instruction sentence (prompt) indicating that it should output, based on the product name, the user information of the user, and the product information database 123, the evaluation axes considered by the user and the scores indicating the importance of the evaluation axes in the category, together with the product name and the user information of the user, into the model. By doing so, the estimation unit 132 estimates the evaluation axes when the user directly selects a product from the input of the product name. The instruction sentence (prompt) is the sentence "Please extract the evaluation axes that are considered important, useful, and effective for me in understanding the characteristics of the input product. Next, with the extracted evaluation axes on the vertical axis and the selected recommended product at the top on the horizontal axis, select products similar on the evaluation axes from the product information database and present them in a table." As a result, the estimation unit 132 can estimate the evaluation axes when the user selects a product directly from the input of the product name.

[0085] Thereafter, the determination unit 133 determines the similarity between products by using the data associated with each product corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis. Then, the providing unit 134 may provide content in which the target product and similar products similar to the target product are arranged in an order corresponding to the similarity between the products.

[0086] 〔7-1. Regarding the processing mode〕 Of the processes described in the above embodiments, all or part of the processes described as being automatically performed can also be performed manually, and conversely, all or part of the processes described as being manually performed can also be automatically performed by known methods. In addition, the processing procedures, specific names, and information including various data and parameters shown in the above text and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

[0087] In addition, each component of each device shown in the drawings is a functional concept and does not necessarily have to be physically configured as shown in the drawings. That is, the specific form of the distribution and integration of each device is not limited to that shown in the drawings, and all or part of it can be functionally or physically distributed and integrated in any unit according to various loads and usage situations.

[0088] In addition, the above-described embodiments can be appropriately combined within a range that does not conflict with the processing content.

[0089] 〔8. Effects〕 As described above, the providing device 1 according to the embodiment includes an estimation unit 132, a determination unit 133, and a providing unit 134. The estimation unit 132 estimates a plurality of evaluation axes and the importance of each evaluation axis when a target product belonging to a specific category is selected, based on the information of the product itself, the user information of the user, and the conversation process regarding a specific category conducted with the user using the conversational service. The determination unit 133 determines the similarity between products using the similarity of the data of each of the plurality of products belonging to a specific category corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis. The providing unit 134 provides content in which the target product and similar products similar to the target product are arranged in an order corresponding to the similarity between products determined by the determination unit 133. Thereby, the providing device 1 can provide a transaction target in consideration of the evaluation axes when the user selects a product.

[0090] Also, in the providing device 1 according to the embodiment, the providing unit 134 provides content in which a target product and similar products are arranged in an order corresponding to at least one of the similarity between products, the advertising fee corresponding to a preset product, the click-through rate, the click unit price obtained by multiplying the advertising fee by the click-through rate, or the value obtained from the advertising fee, the click-through rate, and the lifetime value. Thereby, the providing device 1 can provide the user terminal 10 with the target product and similar products similar to the target product in consideration of the profit of the business operator.

[0091] Also, in the providing device 1 according to the embodiment, the providing unit 134 provides content in which a target product and similar products are arranged such that the closer the data corresponding to the evaluation axis with higher importance is between the target product and a plurality of products, the higher the sorting order. Thereby, the providing device 1 can provide the target product and similar products similar to the target product in consideration of the importance of the evaluation axis when the user selects a product.

[0092] Also, in the providing device 1 according to the embodiment, the providing unit 134 further provides content including a plurality of evaluation axes estimated by the estimation unit 132. Thereby, the providing device 1 can present the evaluation axis when the user selects a product to the user terminal 10.

[0093] Also, in the providing device 1 according to the embodiment, the providing unit 134 provides content including a plurality of evaluation axes rearranged in an order corresponding to the importance of each of the plurality of evaluation axes. Thereby, the providing device 1 can present the evaluation axis when the user selects a product to the user terminal 10 together with the importance.

[0094] Also, in the providing device 1 according to the embodiment, the providing unit 134 displays comparison content beside the evaluation axis that other users consider important among the plurality of evaluation axes. Thereby, the providing device 1 can provide the user with useful information that is not trapped in the user's frame when the user selects a product in a category.

[0095] Also, in the providing device 1 according to the embodiment, the providing unit 134 further changes the importance of each of the plurality of evaluation axes according to the click-through rate to similar products similar to the target product. Thereby, the providing device 1 can find the importance of the evaluation axes according to the actual clicks (accesses) of the user to the similar products.

[0096] 〔9. Hardware Configuration〕 Also, the providing device 1 according to the embodiment described above is realized by, for example, a computer 1000 having a configuration as shown in FIG. 9. Hereinafter, the providing device 1 will be described as an example. FIG. 9 is a hardware configuration diagram showing an example of a computer that realizes the functions of the providing device. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM 1300, an HDD 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a media interface (I / F) 1700.

[0097] The CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400 and controls each part. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program depending on the hardware of the computer 1000, and the like.

[0098] The HDD 1400 stores a program executed by the CPU 1100 and data used by such a program. The communication interface 1500 receives data from other devices via the communication network 500 (corresponding to the network N of the embodiment) and sends it to the CPU 1100, and also sends data generated by the CPU 1100 via the communication network 500 to other devices.

[0099] The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, via the input / output interface 1600. The CPU 1100 acquires data from the input devices via the input / output interface 1600. Also, the CPU 1100 outputs the data generated via the input / output interface 1600 to the output devices.

[0100] The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads such a program from the recording medium 1800 onto the RAM 1200 via the media interface 1700 and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc), a PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

[0101] For example, when the computer 1000 functions as the providing device 1, the CPU 1100 of the computer 1000 realizes the functions of the control unit 13 by executing the program loaded onto the RAM 1200. Also, each data in the storage device of the providing device 1 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be acquired from other devices via a predetermined communication network.

[0102] 〔10. Others〕 As described above in detail some embodiments of the present application with reference to the drawings, these are examples, and the present invention can be implemented in other forms with various modifications and improvements based on the knowledge of those skilled in the art, including the aspects described in the column of the disclosure of the invention.

[0103] In addition, the above-described providing device 1 can have its configuration flexibly changed, such as by calling an external platform or the like through an API (Application Programming Interface) or network computing depending on the function.

[0104] Also, the "section" described in the claims can be read as "means", "circuit", etc. For example, the determination section can be read as a determination means or a determination circuit.

Description of Reference Numerals

[0105] 1 Providing device 10 User terminal 11 Communication section 12 Storage section 13 Control section 121 User information database 122 Estimated information database 123 Product information database 124 Model database 131 Reception section 132 Estimation section 133 Determination section 134 Provision section

Claims

1. Based on information about the product itself, user information of the user, and a conversation process regarding a specific category conducted with the user using a conversational service, a plurality of evaluation axes when a target product belonging to the specific category is selected, and an estimation unit that estimates the importance of each evaluation axis; A determination unit that determines the similarity between products using the similarity of data of each of the plurality of products belonging to the specific category corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis; A providing unit that provides content in which the target product and similar products similar to the target product are arranged in an order according to the similarity between products determined by the determination unit; A providing apparatus characterized by comprising:

2. The providing unit provides content in which the target product and the similar products are arranged in an order according to at least one of the similarity between products, an advertisement fee corresponding to a preset product, a click-through rate, a click unit price obtained by multiplying the advertisement fee by the click-through rate, or a value obtained from the advertisement fee, the click-through rate, and the lifetime value. The providing apparatus according to claim 1, characterized in that:

3. The providing unit provides content in which the target product and the similar products are arranged such that, among the target product and the plurality of products, products in which data corresponding to evaluation axes with higher importance are more similar are ranked higher. The providing apparatus according to claim 1, characterized in that:

4. The providing unit further provides content including the plurality of evaluation axes estimated by the estimation unit. The providing apparatus according to claim 1, characterized in that:

5. The providing unit provides content including the plurality of evaluation axes rearranged in an order according to the importance of each of the plurality of evaluation axes. The providing apparatus according to claim 4, characterized in that:

6. The providing unit provides the content by displaying comparison content beside the evaluation axes that other users consider important among the plurality of evaluation axes. The providing apparatus according to claim 5, characterized in that:

7. The providing unit further changes the importance of each of the plurality of evaluation axes according to the click-through rate to similar products similar to the target product. The providing apparatus according to claim 5, characterized in that:

8. An estimation step of estimating a plurality of evaluation axes and the importance of each evaluation axis when a target product belonging to the specific category is selected, based on information on the product itself, user information of the user, and a conversation process regarding a specific category conducted with the user using a conversational service; A determination step of determining the similarity between products, using the similarity of data of each of the plurality of products belonging to the specific category corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis; A provision step of providing content in which the target product and similar products similar to the target product are arranged in an order corresponding to the similarity between products determined by the determination step; A provision program characterized by causing a computer to execute the above.

9. An estimation procedure of estimating a plurality of evaluation axes and the importance of each evaluation axis when a target product belonging to the specific category is selected, based on information on the product itself, user information of the user, and a conversation process regarding a specific category conducted with the user using a conversational service; A determination procedure of determining the similarity between products, using the similarity of data of each of the plurality of products belonging to the specific category corresponding to each of the plurality of evaluation axes and the importance of each evaluation axis; A provision procedure of providing content in which the target product and similar products similar to the target product are arranged in an order corresponding to the similarity between products determined by the determination procedure; A provision method characterized by including the above.

Images